Re: [PATCH v3 06/25] ASoC: renesas: rz-ssi: Terminate all the DMA transactions

From: Geert Uytterhoeven
Date: Mon Dec 09 2024 - 08:26:40 EST


Hi Claudiu,

On Wed, Nov 13, 2024 at 2:35 PM Claudiu <claudiu.beznea@xxxxxxxxx> wrote:
> From: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>
>
> In case of full duplex the 1st closed stream doesn't benefit from the
> dmaengine_terminate_async(). Call it after the companion stream is
> closed.
>
> Fixes: 4f8cd05a4305 ("ASoC: sh: rz-ssi: Add full duplex support")
> Cc: stable@xxxxxxxxxxxxxxx
> Reviewed-by: Biju Das <biju.das.jz@xxxxxxxxxxxxxx>
> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@xxxxxxxxxxxxxx>

Thanks for your patch!

> Changes in v3:
> - collected tags
> - use proper fixes commit SHA1 and description

I am not sure which one is the correct one: the above, or commit
26ac471c5354583c ("ASoC: sh: rz-ssi: Add SSI DMAC support")...

> --- a/sound/soc/renesas/rz-ssi.c
> +++ b/sound/soc/renesas/rz-ssi.c
> @@ -415,8 +415,12 @@ static int rz_ssi_stop(struct rz_ssi_priv *ssi, struct rz_ssi_stream *strm)
> rz_ssi_reg_mask_setl(ssi, SSICR, SSICR_TEN | SSICR_REN, 0);
>
> /* Cancel all remaining DMA transactions */
> - if (rz_ssi_is_dma_enabled(ssi))
> - dmaengine_terminate_async(strm->dma_ch);
> + if (rz_ssi_is_dma_enabled(ssi)) {
> + if (ssi->playback.dma_ch)
> + dmaengine_terminate_async(ssi->playback.dma_ch);
> + if (ssi->capture.dma_ch)
> + dmaengine_terminate_async(ssi->capture.dma_ch);
> + }

rz_ssi_stop() is called twice: once for capture, and a second time for
playback. How come that doesn't stop both?
Perhaps the checks at the top of rz_ssi_stop() are not correct?
Disclaimer: I am no sound expert, so I may be missing something...

>
> rz_ssi_set_idle(ssi);

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds